Bent-Over Barbell Row

What it trains. The bent-over row is a free-weight horizontal pull done from a hip hinge. Two shoulder actions move the bar:

• Shoulder extension (the upper arm moves back toward and past the trunk): the latissimus dorsi and teres major are the main movers.
• Shoulder horizontal abduction (the upper arm moves back with the elbow out to the side): the posterior deltoid leads, helped by the lats.
• Scapular retraction (the shoulder blades pull together) by the rhomboids and middle and lower trapezius finishes the rep.
• The elbow flexors (biceps brachii, brachialis, brachioradialis) bend the elbow, a secondary role.
• The erector spinae and multifidus work isometrically to hold the hinged torso still.

How far your elbows flare changes the mix. Elbows tucked close to the body make it more of an extension pull for the lats; elbows flared out to the side make it more of a horizontal-abduction pull for the rear delts and upper back. That is a mechanical description of the joint actions, not an EMG finding.

What the evidence says about the lats. A 2026 narrative review of 23 EMG studies on lat activity (Di Fonza et al.) covered the barbell row, bent-over row, seated row, pulldown, pullover and pull-up. Across pulling exercises, it found the humerus-to-trunk relationship may be a key factor in how the load is shared. Among rows, the inverted row produced the highest lat amplitude, but only in one small preliminary study. The authors stress that EMG amplitude is an acute signal, not a training outcome, so treat these comparisons as a starting point.

The low-back cost. Fenwick, Brown and McGill (2009) measured trunk muscle activity and spine load in seven healthy men during three rows:

• The standing bent-over row produced large activation across the whole back, from upper to lower, and the largest lumbar spine load of the three.
• The inverted row produced the highest lat, upper-back and hip-extensor activation with the lowest spine load.
• The one-arm cable row challenged the trunk's resistance to twisting.

Saeterbakken et al. (2015) tested 15 trained men at matched effort. In the bilateral free-weight bent-over row, erector spinae activity was 34% greater than the machine row and 12% greater than the one-arm free-weight row. So the bent-over row trains your back extensors as well as your upper back. That is useful if you want it and a cost if your low back is irritable or already worked hard that day (after deadlifts, for example).

Range of motion. Fischer et al. (2025) had 16 trained men do the prone barbell row through full, upper-half and lower-half range at matched tempo. The lats showed higher mean activity in the upper half, but the effects were medium and inconsistent, and the authors concluded range changes had limited impact on overall activation. For strength, the ACSM 2026 position stand found that training through a complete range of motion improved strength more than partial range. Default to full range: arms long at the bottom, bar to the lower ribs at the top.

Technique cues:

• Grip the bar overhand, a little wider than shoulder width. Hinge at the hips with soft knees until your torso is well inclined (roughly 45° or lower, as far as you can hold a neutral back).
• Brace your trunk and keep your back in its natural curves. Your torso angle should not change during the set.
• Pull the bar toward your lower ribs or upper abdomen. Lead with the elbows and squeeze your shoulder blades together at the top.
• Lower under control to straight arms without letting your chest drop.
• If you have to stand up or swing your hips to move the bar, the load is too heavy for a strict row.

• Intermediate lifters who already own a stable hip hinge (see Deadlifts (bent knee)) and want a heavy free-weight back exercise.
• Anyone who wants back-extensor endurance trained alongside the upper back.
• Lifters who were taught it is a rear-delt exercise and want the current muscle model.

• Healthy backs only. It loaded the lumbar spine the most of the rows Fenwick tested. With current low-back pain, choose a chest-supported row, seated cable row or inverted row (see Inverted Row (Australian Pull-Up)) and get individual advice.
• Small studies. Fenwick had 7 men and Saeterbakken 15, all young and male. They show direction, not exact numbers for you.
• EMG is not growth. No study we found compared muscle growth from the bent-over row with other rows head to head. The muscle roles above come from joint mechanics plus acute EMG.
• Grip strength can limit you before the back does on heavy sets.

Theo's rows turn into a standing shrug-and-heave by the fourth rep.

1. Check the cause: filmed from the side, his torso rises from about 45° to nearly upright as the set goes on. He is using his hips to start each rep, so the load (70 kg) is above what his back can hold in position.
2. Plan for 4 weeks, twice a week:
   • Drop to 55 kg, 3 × 8, with a 1 s pause with the bar touching the lower ribs.
   • The rule: if the torso angle changes, the set ends.
   • Add 2.5 kg when all three sets of 8 stay strict (see Double Progression).
3. Back load: on deadlift days he swaps in a chest-supported row, so his low back isn't worked hard twice in one session.
4. Result: by week 4 he rows 62.5 kg for 3 × 8 with a steady torso and feels it in his lats and mid-back rather than his lower back.

Next lesson: Lat Pulldown. Learn the vertical pull, the row's partner for the lats.

• Primary: latissimus dorsi and teres major (shoulder extension), posterior deltoid (horizontal abduction). Secondary: trapezius, rhomboids, elbow flexors; erector spinae isometric.
• Elbows closer to the body bias the lats; elbows wide bias the rear delts and upper back (joint mechanics).
• Of three rows tested, the standing bent-over row produced the largest lumbar spine load; the inverted row the lowest (Fenwick 2009, n = 7).
• Bilateral free-weight row: erector spinae activity 34% greater than a machine row (Saeterbakken 2015, n = 15).
• Range of motion changed row muscle activity only modestly (Fischer 2025); full range builds more strength (ACSM 2026).